# In the long run, who'd be an aviation reinsurer?

The recent bad run of aviation losses may fail to spoil another good year for underwriters, but Mark Geoghegan wonders why this sector finds it so hard to make a long-term return on capital

The aviation market has been in the news over the month of August as five losses in quick succession have shattered the peace of what was otherwise shaping up to be an exceptional year.

First, on the 2nd of August, an Air France Airbus A330 overran a runway, smashed down a hill and caught fire attempting to land at Toronto's Pearson International Airport in the afternoon. All 297 passengers and 12 crewmembers got out in one piece, but the hull was a total loss, and the compensation bill will be considerable.

Then, just two days later, a Tuninter ATR 72 carrying Italian holidaymakers on a charter flight from Bari, Italy to Tunisia ditched into the sea off the coast of Sicily after losing both engines. According to Airclaims, it would appear that, during impact, the ditching gave some passengers serious injuries and some may have been killed.

Airclaims states that of the 35 passengers and four crew on board, 13 passengers are confirmed to have died, with another two passengers and one crewmember listed as missing. Three crew and eight passengers are believed to have been seriously injured. Tuninter is a subsidiary of Tunisair.

Then, on the 14th of August, a Helios Airways Boeing 737-300 crashed into the side of a mountain near Grammatikos, 40km north of Athens, Greece.

The flight had left Larnaca, Cyprus en route to Prague via Athens, and had lost contact with air-traffic control shortly after the pilot had warned that the aircraft was suffering air-conditioning problems. All 115 passengers and six crew died in the crash.

Barely two days later, a West Caribbean Airways Boeing (MDC) MD-82 crashed in the mountainous area of Machiques, Venezuela. Double engine failure on a charter flight from Panama to Martinique killed all 152 passengers and eight crew on board.

And, as we went to press, reports were coming in of the total loss of a Boeing 737-200 in the jungles of Peru, with the loss of at least 40 lives.

Despite everything - a good year in prospect

However, despite the grisly sequence of events in such a short time, amounting to some $170m in hull losses alone, a glance at the chart shows that (barring major catastrophes), 2005 is actually shaping up to be another good year for underwriters - in fact, the fourth in a row. Steve Doyle, Aviation and Aerospace global practice manager, Aon UK, said: "It is a testament to the work that the industry has carried out over the last decade that the cost of losses is still falling, despite the incidents that have taken place. Accidents are inevitable in this business, but the safety measures implemented are likely to continue to improve safety across the industry." As further evidence of the satisfaction with the overall progress made since 2001, look at this extract from the Lloyd's 2004 annual report: "The loss record for the period 2002 to date has been outstanding, falling to around 50% of the long-term average. For example, 2004 saw the lowest number of commercial jet losses and passenger deaths since records began in 1945." No wonder they were so happy - the Lloyd's aviation market's combined ratio for 2004 was just 79.6%, and a further 7% was available to be released from prior-year reserves, giving a superb final calendar-year technical underwriting profit of 27.4% of capacity. The long-term trend may be for the frequency and severity of losses to have stabilised, and the most recent period may have been particularly benign, but this business is inherently unstable. Let us not forget that it was an appalling year in 2001 that set the current hard market in motion. It is the longer-term trends that provide the industry with its greatest challenges. Consider this: air traffic is set to increase inexorably over the next two decades. According to the projections of Eurocontrol, the European air-traffic control association, even the number of flights in the mature economies of the old continent is set to double over the next 15 years. And since the chances are that this will be matched in the other mature market in North America, and almost certainly outstripped by the developing economies of the Middle East, Asia and South and Central America, we are in for an almost-exponential increase in theoretical collision exposures over the same period. Compensation So, many more planes are going to be flying many more people to many more places much more often. Now for the next factor in the exponentially incremental exposure equation - the inflation of passenger-liability compensation awards. A Swiss Re paper, 'The true value of aviation insurance', quotes a study from the International Union of Aviation Insurers, which looked at the development of 19,000 passenger-compensation awards from 1985 to 2002, and found serious inflation in some areas. Over the 17-year period, the average pay-out to North Americans nearly tripled from$1.1m to $2.9m, equating to an annualised growth rate of 5.87% - this is in fact not bad when inflation and wage growth are factored in. However, it is the Europeans who have made the vastest strides over the same period - their compensation has inflated to a staggering 11.61% annual rate to produce a sevenfold increase, from an average$170,000 to $1.1m. And where Europeans tread, surely Asians and South Americans will not be far behind. The trend over the next 20 years will surely see a further dramatic narrowing of the difference between compensation rates around the globe, as increased prosperity in the developing world translates into higher compensation rates. The recent entry into force of the Montreal convention relating to passenger liability has removed caps on payments, and the updated Rome convention, which has lowered the burden of proof on third-party claims, are only likely to cause the rate of inflation to accelerate across the board. So, we have more planes carrying a more valuable human cargo, probably more exposed to collisions with one another - but add to this the possible effect of the new Airbus A 380. This giant will typically carry 550 passengers, but could hold as many as 840, increasing exposures still further. Experts estimate that at the very least the maximum limit of indemnity bought by large airline should increase for$2bn any one event to $3bn when the A380 comes into service - which represents a 50% increase in financial exposure at the stroke of a pen. Losses - now and in the future And what of the simple contingent liabilities arising from everyday operational delays and causes beyond the operator's control, such as the recent case of thousands of British Airways long-haul travellers being stranded due to problems at an in-flight catering company? A total of 550 delayed customers will cost more per head to house and feed than 400. It all adds up. Clearly, aircraft won't have to be total losses very often for all of this social inflation to start taking its toll on dollar-loss amounts. And it seems that total losses are not the growth area in claims. The excellent Swiss Re paper makes the interesting observation that, while total losses are at the very least stabilising or falling, the number of major partial losses is increasing. The study states that because of greatly improved engine reliability and much more sophisticated cabin-, ground- and collision-avoidance technologies, the frequency of total losses has decreased in recent years; however, partial losses show an inexorable rise. The industry is perhaps a victim of its own sophistication - all the fantastic new equipment is wonderful for avoiding total losses, but is easier to damage and increasingly expensive to repair. Add to this the increasingly complex way in which aircraft are put together with a greater use of composite materials, and you're looking at an expensive cocktail. So the future holds more increasingly complex and expensive aircraft flying more increasingly expensive passengers - what else is there to worry about? Well, the threat of terrorism is no less prevalent now than five years ago and, although few major losses have been yet suffered by the industry due to weather, we should not discount the increase in frequency and severity of natural-catastrophe perils noted by the reinsurance community in the rest of the world. Swiss Re estimate that approximately$7bn worth of aircraft are on the ground at any given moment at Tokyo's Haneda airport, which is exposed to both severe windstorm and earthquake risks.

But enough of this doom and gloom, and theoretical exposures - what of the state of the aviation insurance market today? Let's get back to Lloyd's - the insurance market's 2004 annual report talks of "downward pressure on prices".

Results of market softening

After three great years, no wonder the market is softening right now - who wouldn't want to get their hands on 'spectacular' returns like these?

The problem is that the relatively small size of the market means that additional capacity coming from other classes makes a large splash almost immediately.

As an example, another Swiss Re study of 2002 estimated that "if a mere 0.1% of all non-life underwriting capacity flows into the aviation market, this may correspond to a 25% increase in capacity" - no wonder rates are notoriously unstable in this class. And the long-term underwriting result has been scandalously poor. Even Lloyd's results look a little less superb when considered in the context of record benign-loss experience.

If loss frequency and severity have been half the average, current premium levels can't be anything more than the absolute minimum required for the job. What would the result have been in an average year? One presumes a combined ratio of around 100%. So absolutely nothing is currently left in the kitty for above-average years or catastrophic losses?

With total hull sums insured approaching $600bn, it seems staggering that for the 1992-2002 10-year period, insurers managed to take in only an average of$1.4bn a year in premiums.

The result speaks for itself - Swiss Re found that over the 1988-2002 period, the industry lived way beyond its means, racking up cumulative underwriting deficit of some $2.3bn. Premiums always rose after significantly poor periods of results, but never sufficiently to pay back what was owed. The class lost money in 1988-93, hardened and made money in 1994 and 1995 - only to soften and be losing again by 1997, without having taken a big chunk out of its cumulative deficit. More up-to-date figures are not available for the aviation class as a whole, but at 2001, the deficit stood at over$6bn.

Despite three-and-a-half good years, it is extremely unlikely that this sum is to be made back before softening occurs and results begin to slide again. Who'd be an aviation (re)insurer, indeed?